4  Yong-of-the-year (YOY) Brook Trout (Salvelinus fontinalis) Body Size Model

5 Background

  • The Ecology Section at the USGS Conte Laboratory has studied brook trout in West Brook, a small 3rd order stream in Whately MA, since 1997.
  • They have observed body size variation in Young-of-the-Year (YOY) brook trout (Salvelinus fontinalis) in the fall. ( Photo above )
  • Knowing the mechanisms behind the body size variation are important because a larger body size increase the probability of overwinter survival and age at first reproduction. There might be an underlying strategy that allows the parents to be the drivers of the population.
  • What is causing the observed body size variation in YOY brook trout?

5.1 Questions

Although there are a lot of variables like, food availability, habitat, flow and temperature all influenceing body size. I want to look at it through the lenses of two variables.

  1. What effect does growth rate have on brook trout YOY body size?
  2. What effect does emergence date ( birth ) have on brook trout YOY body size?

5.2 Study Structure

  • Conceptual Models

  • Otolith Growth Rate Methodology

  • Stream Data

6 Conceptual Model

6.1 Simulated Data

[1] "Same Emergence/Diff Growth"

6.2 Simulated Data effect on Body Size

7 Stream Data

7.1 Avery Brook Size Distribution vs Emergence Date (2022-07-21)

7.2 Conclusions

  1. Brook Trout Length and Growth Rate are highly correlated.
  2. Need an independent measure of growth rate.
  3. Use otolith ring spacing to estimate daily growth rate.

8 Back calculation of fish size from otolith size

8.1 Steps

  1. Determine the relationship between brook trout fork length and otolith size.
  2. Access the daily body size estimates from ring spacing measurements.
  3. Apply methodology to one stream from one year.

8.2 Otolith Radius vs Body Size

  1. There is a positive relationship between otolith radius and fork length.
    1. R Squared = 0.92
  2. We conducted a mixed effects model analysis to determine whether the sampled stream and year could explain the variance. Unfortunately, the random effects only contributed a minimal amount to the overall variance.

8.3 Ring Spacing Methodology

  1. A sagittal otolith is divided into four quadrants.
    1. Posterior-Ventral
    2. Posterior-Dorsal
    3. Anterior-Ventral
    4. Anterior-Dorsal
  2. All Counting was done in the Posterior-Dorsal Quadrant.

  1. Ring spacing is measured along a 45 degree angle in the Posterior-Dorsal Quadrant
    • Starting at the Posterior most primordia of the primordium.
    • Ending at the edge of the Otolith.
  2. This ensures consistent measurements among samples.

Posterior-Dorsal Quadrant with Counting Line
  1. The distance between every consecutive ring was measured starting at the edge and ending at the posterior primordium.
  2. The distances between the consecutive rings were used to calculate the otolith radius when the ring was formed.
  3. The otolith radius is then run through a linear equation to calculate the Fishes Fork Length for a given rings size.

Zoomed in Posterior-Dorsal Quadrant.

8.4 Error Checking

8.4.1 Ring Measuring Error

  1. Check for reader bias.
  2. Measure ring spacing and calculate estimated body size for Ten otolith blindly.
  3. Compare the size estimates and calculate the percent error in them.
  4. Percent Error should be below 10%.
  5. Two out of Ten percent error over 10%.

9 Application to Stream Data ( Avery Brook 2022)

9.0.1 Avery Brook Growth Rate Data Raw

The next step after evaluation emergence and growth rates effects on brook trout YOY size distribution will be to look at the effects of discharge.

9.0.1.1 Daily Growth Rate

9.0.1.2 Average Growth Comparisons ( Filler Tab )

Code 
summary( lm( forkLengthFinal ~ meanGrowth  , data = emergenceDateModel ))

Call:
lm(formula = forkLengthFinal ~ meanGrowth, data = emergenceDateModel)

Residuals:
     Min       1Q   Median       3Q      Max 
-18.1364  -5.1063  -0.3208   5.0588  20.0254 

Coefficients:
            Estimate Std. Error t value Pr(>|t|)    
(Intercept)    3.705      2.987    1.24     0.22    
meanGrowth   131.773     10.792   12.21   <2e-16 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 8.069 on 61 degrees of freedom
Multiple R-squared:  0.7096,    Adjusted R-squared:  0.7049 
F-statistic: 149.1 on 1 and 61 DF,  p-value: < 2.2e-16
Code 
summary(  lm( forkLengthFinal ~ emergenceSizeDOY , data = emergenceDateModel ))

Call:
lm(formula = forkLengthFinal ~ emergenceSizeDOY, data = emergenceDateModel)

Residuals:
    Min      1Q  Median      3Q     Max 
-22.405 -11.065  -1.616  11.459  39.226 

Coefficients:
                 Estimate Std. Error t value Pr(>|t|)   
(Intercept)      -10.6765    18.0312  -0.592  0.55596   
emergenceSizeDOY   0.4316     0.1591   2.713  0.00866 **
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 14.14 on 61 degrees of freedom
Multiple R-squared:  0.1077,    Adjusted R-squared:  0.09303 
F-statistic:  7.36 on 1 and 61 DF,  p-value: 0.008657
Code 
summary( lm( forkLengthFinal ~ emergenceSizeDOY , data = emergenceDateModel ))

Call:
lm(formula = forkLengthFinal ~ emergenceSizeDOY, data = emergenceDateModel)

Residuals:
    Min      1Q  Median      3Q     Max 
-22.405 -11.065  -1.616  11.459  39.226 

Coefficients:
                 Estimate Std. Error t value Pr(>|t|)   
(Intercept)      -10.6765    18.0312  -0.592  0.55596   
emergenceSizeDOY   0.4316     0.1591   2.713  0.00866 **
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Residual standard error: 14.14 on 61 degrees of freedom
Multiple R-squared:  0.1077,    Adjusted R-squared:  0.09303 
F-statistic:  7.36 on 1 and 61 DF,  p-value: 0.008657